The complication of GVHD currently prohibits the application of bone marrow transplantation across extensive HLA barriers. T cell depletion of allogeneic marrow effectively prevents GVHD, but is associated with increased risks of leukemic relapse and of engraftment failure. We have shown that GVHD-promoting activity can be separated from the graft-vs- leukemia (GVL) and engraftment-promoting effects of host-reactive T cells, and the goal of this proposal is to further explore strategies for achieving this separation. These are: 1) Prophylaxis of mice with a short peri-transplant course of IL-2 inhibits GVHD while preserving engraftment-promoting and GVL effects of donor T cells. IL-2 selectively inhibits CD4 activity, and preserves GVL effects of CD8+ cells. Furthermore, while inhibiting CD4-dependent GVHD=mediated GVL against a class II MHC-positive pro-monocytic leukemia. Therefore, separate activities of CD4+ T cells can mediate GVHD or GVL effects, and these can be dissociated by IL-2 treatment. We will attempt to distinguish these activities and to determine which ones are inhibited by IL-2 treatment by evaluating the role of cytolytic cells and of various cytokines in CD4-mediated GVHD and in CD4- and CD8-mediated GVL effects in the class II-positive promonocytic leukemia model; 2) CD8+ T cells can mediate CD4-independent GVL effects, but are largely CD4-dependent for their capacity to cause severe acute GVHD in fully histoincompatible strain combinations. In fact, when donor CD8+ cells are given without CD4+ cells, the severity of GVHD directed against minor histoincompatibilities alone is at least as severe as that directed against the same minor antigens in association with allogeneic MHC (including class I) antigens. This failure of MHC disparity to increase the severity of CD9-mediated GVHD suggests that administration of CD8+ T cells without CD4+ cells might permit HLA-mismatched BMT to be performed with no greater incidence of GVHD than is observed for HLA-matched sibling transplants. To evaluate the potential of this approach, we will attempt to delineate the genetic factors and host cell populations that influence the capacity of CD8+ T cells given without CD4+ cells to cause GVHD, and will determine the requirement for each T cell subset to achieve optimal alloengraftment in the setting of various genetic disparities. Finally, we will: 3) evaluate the possibility that cure of chronic lymphocytic leukemia could be achieved without ablative conditioning by using allogeneic T cell infusion to established mixed chimeras. Donor T cells administered to established mixed allogeneic chimeras prepared with a mild, non- myeloablative conditioning regimen lymphohematopoietic GVH reactions without causing GVHD. These lymphohematopoietic GVH reactions might be associated with GVL effects. This approach could be applicable in elderly patients with chronic lymphocytic leukemia, who are not eligible for conventional BMT because of their high incidence of GVHD. Together, our studies may lead to new clinical strategies for preserving potentially potent GVL effects of MHC-reactive T cells while minimizing GVHD.